plateform stringclasses 1
value | repo_name stringlengths 13 113 | name stringlengths 3 74 | ext stringclasses 1
value | path stringlengths 12 229 | size int64 23 843k | source_encoding stringclasses 9
values | md5 stringlengths 32 32 | text stringlengths 23 843k |
|---|---|---|---|---|---|---|---|---|
github | domingomery/Balu-master | Bmv_bundleafin.m | .m | Balu-master/MultiView/Bmv_bundleafin.m | 2,780 | utf_8 | 000ec1b973b8d1892d05949024965985 | % [Xs,Ps,xs] = Bmv_bundleafin(x)
%
% Toolbox: Balu
%
% Bundle Adjustment Projective reconstruction
% using the factorization algorithm
%
% input:
% x projected 2D points as 3 x n x m matrix (homogeneous)
% with n number of 3D points
% m number of views
%
% output:
% Xs esti... |
github | domingomery/Balu-master | Bmv_homographySIFT.m | .m | Balu-master/MultiView/Bmv_homographySIFT.m | 2,393 | utf_8 | 562ec95f702f494c74e8bef9fafd01ab | % [Ibs,H] = Bmv_homographySIFT(Ia,Ib,show)
%
% Toolbox: Balu
%
% Homography between images Ia and Ib using RANSAC of SIFT points
% Ibs is transformed image Ib.
% size(Ia) = size(Ibs)
% If ma and mb are the homogeneus coordinates of points in images Ia and Ib:
% ma = [xa ya 1]'; mb = [xb yb 1]';
% ... |
github | domingomery/Balu-master | Bmv_lines2point.m | .m | Balu-master/MultiView/Bmv_lines2point.m | 566 | utf_8 | dcb75873c3f972328fca86df4fe397f8 | % function m = Bmv_lines2point(l1,l2)
%
% Toolbox Balu
%
% 2D point m computed as the intersection of two 2D lines (l1 and l2).
%
% m = Bmv_lines2point(l1,l2) returns the 2D point m defined
% as the intersection of lines l1 and l2.
% m, l1, l2 are 3x1 homogeneous vectors. The result
% m is given as m ... |
github | domingomery/Balu-master | Bmv_homographyRANSAC.m | .m | Balu-master/MultiView/Bmv_homographyRANSAC.m | 1,748 | utf_8 | a15f886452bd42a43f313955d760d6d5 | % H = Bmv_homographyRANSAC(m1,m2)
%
% Toolbox: Balu
%
% Estimation of Homography Matrix using RANSAC.
%
% m1 and m2 are n corresponding points in two views (m1(:,k) and m2(:,k)
% are the k-th corresponding points (for k=1..n) stored as homogeneous
% coordinates.
%
% Example:
% m1 = [rand(2,20);ones... |
github | domingomery/Balu-master | Bmv_trifocal.m | .m | Balu-master/MultiView/Bmv_trifocal.m | 1,150 | utf_8 | 6d01fc87932bdc32c6f9362bba8fc6f8 | % F = Bmv_trifocal(A,B,C)
%
% Toolbox: Balu
%
% Trifocal tensors.
%
% T = trifocal(A,B,method) returns the trifocal tensors from
% 3x4 projection matrices A, B and C according of thre
% views. T is a 3x3x3 array
%
% The method can be found in:
%
% R. Hartley and A. Zisserman. Multiple View... |
github | domingomery/Balu-master | Bmv_fundamentalSIFT.m | .m | Balu-master/MultiView/Bmv_fundamentalSIFT.m | 1,621 | utf_8 | b7ce436f585d807e6f09ed680f4025db | % F = Bmv_fundamentalSIFT(I1,I2)
%
% Toolbox: Balu
%
% Estimation of Fundamental Matrix from two images using SIFT points.
%
% This function requires VLFeat Toolbox from (www.vlfeat.org).
%
% I1 and I2 are the stereo images.
% F is the Fundamental matrix.
%
% Example:
% I1 = imread('testimg5.jpg');... |
github | domingomery/Balu-master | Bmv_fundamentalRANSAC.m | .m | Balu-master/MultiView/Bmv_fundamentalRANSAC.m | 2,397 | utf_8 | 33c5a162473524b429e5fafea822b5cd | % [F,inliers] = Bmv_fundamentalRANSAC(m1,m2)
%
% Toolbox: Balu
%
% Estimation of Fundamental Matrix using SVD decomposition (Hartley,
% p.281) with det(F) = 0, norm(F(:)) = 1.
%
% m1 and m2 are n corresponding points in two views (m1(:,k) and m2(:,k)
% are the k-th corresponding points (for k=1..n) stored... |
github | domingomery/Balu-master | Bmv_bundleproj.m | .m | Balu-master/MultiView/Bmv_bundleproj.m | 3,389 | utf_8 | bf9685aa821b503c48d547ac1b2ecfda | % [Xs,Ps,xs] = Bmv_bundleproj(x)
%
% Toolbox: Balu
%
% Bundle Adjustment Projective reconstruction
% using the factorization algorithm
%
% input:
% x projected 2D points as 3 x n x m matrix (homogeneous)
% with n number of 3D points
% m number of views
%
% output:
% Xs esti... |
github | domingomery/Balu-master | Bmv_reproj3.m | .m | Balu-master/MultiView/Bmv_reproj3.m | 1,674 | utf_8 | bc1950395fecb7078388a830bf99151b | % function m3s = Bmv_reproj3(m1,m2,T,method)
%
% Toolbox Balu:
%
% Reprojection of point m3 from m1, m2 and trifocal tensors
%
% m3s = reproj3(m1,m2,T) returns the reprojection of m3 from
% corresponding points m1 and m2 in image 1 and 2 respectivelly
% using trifocal tensors.
% method = 1 uses the fi... |
github | domingomery/Balu-master | Bmv_matrixp.m | .m | Balu-master/MultiView/Bmv_matrixp.m | 592 | utf_8 | 1524ec9a89203f0447dfdf888bbb0457 | % function P = Bmv_matrixp(f)
%
% Toolbox Balu
%
% Perspective proyection matrix 3D->2D.
%
% It returns the 3x4 perspective proyection matrix
% depending on focal distance f.
%
% Bmv_matrixp(f) is equal to
% [f 0 0 0
% 0 f 0 0
% 0 0 1 0]
%
% Example:
%
% f = 10; % foca... |
github | domingomery/Balu-master | Bmv_reco3dn.m | .m | Balu-master/MultiView/Bmv_reco3dn.m | 2,197 | utf_8 | 9fd52416d6dfdd8083d497d6930fa513 | % [M,err,ms] = Bmv_reco3dn(m,P)
%
% Toolbox: Balu
%
% 3D reconstruction from n corresponding points
%
% It returns a 3D point M that fullfils
% the following projective equations:
%
% lambda1*m1 = P1*M
% lambda2*m2 = P2*M
% :
% where mk = m(:,k) are the 2D projection points of 3D point M
... |
github | domingomery/Balu-master | Bmv_matrixr2d.m | .m | Balu-master/MultiView/Bmv_matrixr2d.m | 457 | utf_8 | 2b34df217006a08a31efd47f1b8137cb | % function R = Bmv_matrixr2d(theta)
%
% Toolbox Balu:
%
% 2D rotation matrix.
%
% It returns the 2D rotation matrix given by a rotation
% theta (given in radians):
%
% R = [ cos(theta) -sin(theta)
% sin(theta) cos(theta)];
%
%
% Example:
%
% R = matrixr2d(pi/3)
%
%
% (c) D.Mery, PUC-DCC,... |
github | domingomery/Balu-master | Bmv_epiplot.m | .m | Balu-master/MultiView/Bmv_epiplot.m | 1,045 | utf_8 | 696f50832b76f47337c202efccfc07ef | % ell = Bmv_epiplot(F,m1)
%
% Toolbox: Balu
%
% Plot of epipolar line.
%
% The epipolar line is ell = F*m1. F is the Fundamental Matrix and m1 is
% a point image 1 in homogeneous coordinates.
%
% Example:
% I1 = imread('testimg5.jpg'); % Image 1
% figure(1);imshow(I1); hold on
% ... |
github | domingomery/Balu-master | Bmv_reco3d2.m | .m | Balu-master/MultiView/Bmv_reco3d2.m | 1,469 | utf_8 | cfa471f29d5349ec9c8ffd3146be685d | % M = Bmv_reco3d2(m1,m2,A,B)
%
% Toolbox: Balu
%
% 3D reconstruction from n corresponding points
%
% It returns a 3D point M that fullfils
% the following projective equations:
%
% lambda1*m1 = A*M
% lambda2*m2 = B*M
% :
% where mk are the 2D projection points of 3D point M
% in image... |
github | domingomery/Balu-master | Bmv_fundamental.m | .m | Balu-master/MultiView/Bmv_fundamental.m | 2,358 | utf_8 | acb61f023a5a9a198991592a91216de6 | % F = Bmv_fundamental(A,B,method)
%
% Toolbox: Balu
%
% Fundamental matrix from projection matrices.
%
% F = fundamental(A,B,method) returns the 3x3 fundamental matrix from
% 3x4 projection matrices A and B according to the following
% methods:
%
% method = 'tensor' : uses bifocal tensors with canoni... |
github | domingomery/Balu-master | Bmv_homographyRSIFT.m | .m | Balu-master/MultiView/Bmv_homographyRSIFT.m | 3,401 | utf_8 | e1488cb11e2de734eff13d349ea832b3 | % [Ibs,H] = Bmv_homographySIFT(Ia,Ib,Ra,Rb,show)
%
% Toolbox: Balu
%
% Homography between images Ia and Ib using RANSAC of SIFT points
% Ibs is transformed image Ib.
% Ra and Rb are binary images that indicate where the keypoints are
% valid, ie, SIFT descriptors in pixels of Ia (or Ib) where Ra = 0 (or
% ... |
github | domingomery/Balu-master | Bmv_fundamentalSVD.m | .m | Balu-master/MultiView/Bmv_fundamentalSVD.m | 1,314 | utf_8 | 49d06ed201d3b98efdd571b49881a8ac | % F = Bmv_fundamentalSVD(m1,m2)
%
% Toolbox: Balu
%
% Estimation of Fundamental Matrix using SVD decomposition (Hartley,
% p.281) with det(F) = 0, norm(F(:)) = 1.
%
% m1 and m2 are n corresponding points in two views (m1(:,k) and m2(:,k)
% are the k-th corresponding points (for k=1..n) stored as homogeneo... |
github | domingomery/Balu-master | Bmv_matchSIFT.m | .m | Balu-master/MultiView/Bmv_matchSIFT.m | 5,470 | utf_8 | 1639b4ef792348af7c0ec4d62211b6f0 | % [f1,d1,f2,d2,scores] = Bmv_matchSIFT(I1,I2,method,show)
%
% Toolbox: Balu
%
% Matching points between two images I1 and I2 using SIFT points.
%
% This function requires VLFeat Toolbox from (www.vlfeat.org).
%
% method = 1 is for vl_ubcmatch method
% method = 2 is for vl_ubcmatch plus RANSAC with homograph... |
github | domingomery/Balu-master | Bmv_points2line.m | .m | Balu-master/MultiView/Bmv_points2line.m | 787 | utf_8 | 5f0d05fdb0316caf9dbfbd5f7136e680 | % function l = Bmv_points2line(m1,m2)
%
% Toolbox Balu
%
% 2D line l that contains two 2D points (m1 and m2).
%
% l = points2line(m1,m2) returns the 2D line l defined as the line that
% contains the 2D points m1 and m2. l, m1, m2 are 3x1 homogeneous
% vectors. Points m1 and m2 can be 2x1 vectors. The ... |
github | domingomery/Balu-master | Bmv_epipoles.m | .m | Balu-master/MultiView/Bmv_epipoles.m | 1,080 | utf_8 | 9be3d28a930a999ab8718efe256601e4 | % [e1,e2] = Bmv_epipoles(F)
%
% Toolbox: Balu
%
% Epipoles of a two-view system from fundamental matrix.
%
% [e1,e2] = epipoles(F) returns:
% e1: epipole in view 1
% e2: epipole in view 2
% where e1 and e2 are 3x1 homogeneous vectors.
% e1(3) = e2(3) is 1.
%
% Example:
% A = rand(3,4);... |
github | domingomery/Balu-master | Bmv_guiproy2D.m | .m | Balu-master/MultiView/Bmv_guiproy2D.m | 12,430 | utf_8 | 24008ea26d0c77d8fe779936f797cb50 | function varargout = Bmv_guiproy2D(varargin)
% BMV_GUIPROY2D M-file for Bmv_guiproy2D.fig
% BMV_GUIPROY2D, by itself, creates a new BMV_GUIPROY2D or raises the existing
% singleton*.
%
% H = BMV_GUIPROY2D returns the handle to a new BMV_GUIPROY2D or the handle to
% the existing singleton*.
%
... |
github | domingomery/Balu-master | Bmv_matrixr3d.m | .m | Balu-master/MultiView/Bmv_matrixr3d.m | 1,161 | utf_8 | f0c59ada44a9c9867aa200c14898dc0a | % function R = Bmv_matrixr3d(wx,wy,wz)
%
% Toolbox Balu:
%
% 3D rotation matrix.
%
% It returns the 3D rotation matrix given by a rotation
% arround z, y and x axes where the rotation angles are wz, wy, and
% wx respectively. The angles are given in radians.
%
% R = Bmv_matrixr3d(wx,wy,wz) is equal t... |
github | domingomery/Balu-master | Bmv_trifocalSVD.m | .m | Balu-master/MultiView/Bmv_trifocalSVD.m | 2,831 | utf_8 | 51a1384cd167a12f31790f8ba1a087b1 | % F = Bmv_trifocalSVD(m1,m2,m3)
%
% Toolbox: Balu
%
% Estimation of Trifocal Tensors using SVD decomposition.
%
% m1, m2 and m3 are n corresponding points in 3 views m1(:,k), m2(:,k)
% and m3(:,k) are the k-th corresponding points (for k=1..n) stored as
% homogeneous coordinates.
%
% Example:
% A ... |
github | domingomery/Balu-master | Bmv_homographySVD.m | .m | Balu-master/MultiView/Bmv_homographySVD.m | 1,301 | utf_8 | 7f492d0184a633021f2c37322dc52fec | % H = Bmv_homographySVD(m1,m2)
%
% Toolbox: Balu
%
% Estimation of Homography Matrix using SVD decomposition.
%
% m1 and m2 are n corresponding points in two views (m1(:,k) and m2(:,k)
% are the k-th corresponding points (for k=1..n) stored as homogeneous
% coordinates.
%
% Example:
% m1 = [rand(2,... |
github | domingomery/Balu-master | Bfa_corrsearch.m | .m | Balu-master/FeatureAnalysis/Bfa_corrsearch.m | 2,411 | utf_8 | baa130a90e717529d30be08d94c8ee56 | % [per,R] = Bfa_corrsearch(x,y,method,v,show)
%
% Toolbox: Balu
% Error estimation of linear or quadratic model
% that minimizes the norm between measured and modeled
% output. The error is estimated using cross-validation.
%
% x: measured input (nxm : n samples and m variables)
% y: measured out... |
github | domingomery/Balu-master | Bfa_sp100.m | .m | Balu-master/FeatureAnalysis/Bfa_sp100.m | 865 | utf_8 | 25a099904f183778e03c12622407ffb3 | % Sp = Bfa_sp100(X,d)
%
% Toolbox: Balu
% Especificty at Sensibility = 100%.
% X features matrix. X(i,j) is the feature j of sample i.
% d vector that indicates the ideal classification of the samples
%
% See also Bfs_sfs, Bfa_fisher
%
% (c) D.Mery, PUC-DCC, 2011
% http://dmery.ing.puc.cl
function... |
github | domingomery/Balu-master | Bfa_bestcorrn.m | .m | Balu-master/FeatureAnalysis/Bfa_bestcorrn.m | 3,261 | utf_8 | f8d7d91ac3c4fbd4403db8653d9e8509 | % [selec,cx,B] = Bfa_bestcorrn(X,y)
%
% Toolbox: Balu
% Search of the variables of X that best correlate with y.
%
% Three models are obtained:
% y = a1*z1 + a0 > see figure 1
% y = a1*z1 + a2*z2 + a0 > see figure 2
% y = a1*z1 + a2*z2 + a3*z3 + a0 > see figure 3
%
% sel... |
github | domingomery/Balu-master | Bfa_gmean.m | .m | Balu-master/FeatureAnalysis/Bfa_gmean.m | 1,042 | utf_8 | ab651dd6c48e5f24ff2857ea689f0a05 | % y = Bfa_gmean(X,d,op)
%
% Toolbox: Balu
% op = 1: sqrt (Specificty * Sensibility)
% op = 2: sqrt (Precision * Recall)
% X features matrix. X(i,j) is the feature j of sample i.
% d vector that indicates the ideal classification of the samples
%
% See also Bfs_sfs, Bfa_fisher
%
% (c) D.Mery, PUC-D... |
github | domingomery/Balu-master | Bfa_miparzen2.m | .m | Balu-master/FeatureAnalysis/Bfa_miparzen2.m | 1,519 | utf_8 | f87f513e01481890f7539ff91bcf021a | % Mutual Information using Parzen windows for two variables
% NOTE:
% The pdf's are estimated using Kernel Density Estimations programs
% kde.m and kde2d.m after Botev et al. (2010) implemented by Botev.
% These files are in Balu directory 'Feature Analysis' as Bfa_kde and
% Bfs_kde2d. They can also be do... |
github | domingomery/Balu-master | Bfa_bestcorr.m | .m | Balu-master/FeatureAnalysis/Bfa_bestcorr.m | 1,687 | utf_8 | ec4092aab4247e3252a0fea5e1e67c9a | % [selec,cx,a] = Bfa_bestcorr(X,y)
%
% Toolbox: Balu
% Search of the variables of X that best correlate with y.
%
% selec is the number of the selected variables
% cx is the correlation coefficient
% a is the parameters vector of the model
% selec and cx are sorted (eg, selec(i) is the number o... |
github | domingomery/Balu-master | Bfa_kde.m | .m | Balu-master/FeatureAnalysis/Bfa_kde.m | 5,546 | utf_8 | cbd93bee97cb218abfdfeb2f772a6dd1 | % NOTE:
% This file corresponds to kde.m implemented by Zdravko Botev.
% It can also be downloaded from www.mathwork.com
% (c) Zdravko Botev. All rights reserved.
%
function [bandwidth,density,xmesh,cdf]=Bfa_kde(data,n,MIN,MAX)
% Reliable and extremely fast kernel density estimator for one-dimensional d... |
github | domingomery/Balu-master | Bfa_dXi2.m | .m | Balu-master/FeatureAnalysis/Bfa_dXi2.m | 372 | utf_8 | c633d33e99be431635750a676750abe6 | % d = Bfa_dXi2(X,Y)
%
% Toolbox: Balu
%
% Xi^2 distance between two vectors X and Y
%
% Example:
% X = (1:50)';
% Y = X + randn(50,1);
% d = Bfa_dXi2(X,Y)
%
%
% (c) D.Mery, PUC-DCC, 2010
% http://dmery.ing.puc.cl
function d = Bfa_dXi2(X,Y)
X = double(X);
Y = double(Y);
s = X+Y;
... |
github | domingomery/Balu-master | Bfa_jfisher.m | .m | Balu-master/FeatureAnalysis/Bfa_jfisher.m | 1,067 | utf_8 | 0e3c3c7cbd6844b89a699f75f0f27a07 | % J = Bfa_jfisher(X,d,p)
%
% Toolbox: Balu
% Fisher objective function J.
% X features matrix. X(i,j) is the feature j of sample i.
% d vector that indicates the ideal classification of the samples
% p a priori probability of each class
%
% See also Bfs_sfs.
%
% (c) D.Mery, PUC-DCC, 2011
% http:/... |
github | domingomery/Balu-master | Bfa_sqcorrcoef.m | .m | Balu-master/FeatureAnalysis/Bfa_sqcorrcoef.m | 445 | utf_8 | a5e1a06cdd3a197a250a42fca12bfb22 | % sc = Bfa_sqcorrcoef(x,y)
%
% Toolbox: Balu
% Squared-correlation coefficient between two random vector x and y
%
% Wei, H.-L. & Billings, S. Feature Subset Selection and Ranking for
% Data Dimensionality Reduction Pattern Analysis and Machine
% Intelligence, IEEE Transactions on, 2007, 29, 162-166
%
% D.... |
github | domingomery/Balu-master | Bfa_vecsimilarity.m | .m | Balu-master/FeatureAnalysis/Bfa_vecsimilarity.m | 916 | utf_8 | 1f4a3da625633dc78a344e7ceaa7a114 | % [rk,j] = Bvecsimilarity(vq,v)
%
% Toolbox: Balu
%
% Normalized scalar product (cosine angle).
%
% vq: query vector
% v : family of vectors
% This function serach the minimal distance between vq and all vector
% in v.
% rk are the sorted scalar product.
% j are the sorted indices.
%
% ... |
github | domingomery/Balu-master | Bfa_kde2d.m | .m | Balu-master/FeatureAnalysis/Bfa_kde2d.m | 7,694 | utf_8 | 2bc1167be519ee41eda99f47d1616a18 | % NOTE:
% This file corresponds to kde2d.m implemented by Zdravko Botev.
% It can also be downloaded from www.mathwork.com
% (c) Zdravko Botev. All rights reserved.
%
% fast and accurate state-of-the-art
% bivariate kernel density estimator
% with diagonal bandwidth matrix.
% The kernel is assumed to ... |
github | domingomery/Balu-master | Bcl_ensemble.m | .m | Balu-master/Classification/Bcl_ensemble.m | 6,284 | utf_8 | 23900419d43d4e31cf3c69a539b6ea02 | % ds = Bcl_ensemble(X,d,Xt,options) Training & Testing together
% options = Bcl_ensemble(X,d,options) Training only
% ds = Bcl_ensemble(Xt,options) Testing only
%
% Toolbox: Balu
% Design and test an ensemble of n classifiers.
%
% Design data:
% X is a matrix with features (columns)
% ... |
github | domingomery/Balu-master | Bcl_AdaBoostM1.m | .m | Balu-master/Classification/Bcl_AdaBoostM1.m | 9,141 | utf_8 | 2125e39cd7d3ea25f028c312d707c1eb | % ds = Bcl_AdaBoostM1(X,d,Xt,options) Training & Testing together
% options = Bcl_AdaBoostM1(X,d,options) Training only
% ds = Bcl_AdaBoostM1(Xt,options) Testing only
%
% Toolbox: Balu
% AdaBoost M1 classifier for two classes
%
% Design data:
% X is a matrix with features (column... |
github | domingomery/Balu-master | Bcl_libsvm.m | .m | Balu-master/Classification/Bcl_libsvm.m | 4,292 | utf_8 | c5782c14193c1659a97b8bbfaa7de206 | % ds = Bcl_svm(X,d,Xt,options) Training & Testing together
% options = Bcl_svm(X,d,options) Training only
% ds = Bcl_svm(Xt,options) Testing only
%
% Toolbox: Balu
% Support Vector Machine approach using the LIBSVM(*).
%
% Design data:
% X is a matrix with features (columns)
% d is... |
github | domingomery/Balu-master | Bcl_weakc.m | .m | Balu-master/Classification/Bcl_weakc.m | 2,884 | utf_8 | 87570ff35169e967569a77b3b5316e71 | % ds = Bcl_weakc(X,d,Xt,options) Training & Testing together
% options = Bcl_weakc(X,d,options) Training only
% ds = Bcl_weakc(Xt,options) Testing only
%
% Toolbox: Balu
% Weak classifier for one feature X using Otsu method.
% Design data:
% X is a column vector with only one feature
% ... |
github | domingomery/Balu-master | Bcl_adaboost.m | .m | Balu-master/Classification/Bcl_adaboost.m | 4,292 | utf_8 | ec04bb72b67429fabf72e5ebe9153ff8 | % ds = Bcl_adaboost(X,d,Xt,options) Training & Testing together
% options = Bcl_adaboost(X,d,options) Training only
% ds = Bcl_adaboost(Xt,options) Testing only
%
% Toolbox: Balu
% AdaBoost.M2 classifier.
%
% Design data:
% X is a matrix with features (columns)
% d is the ... |
github | domingomery/Balu-master | Bcl_bayes2.m | .m | Balu-master/Classification/Bcl_bayes2.m | 5,580 | utf_8 | 811aa6c94ebe62f3bae6f37f73cedef8 | % ds = Bcl_bayes2(X,d,Xt,options) Training & Testing together
% options = Bcl_bayes2(X,d,options) Training only
% ds = Bcl_bayes2(Xt,options) Testing only
%
% Toolbox: Balu
% Bayes classifier for ONLY two features and two classes
%
% Design data:
% X is a matrix with features (co... |
github | domingomery/Balu-master | Bcl_tree.m | .m | Balu-master/Classification/Bcl_tree.m | 3,669 | utf_8 | aee9da403c47541c3071f93b7a6b4537 | % ds = Bcl_tree(X,d,Xt,options) Training & Testing together
% options = Bcl_tree(X,d,options) Training only
% ds = Bcl_tree(Xt,options) Testing only
%
% Toolbox: Balu
% Classifier using a tree algorithm
%
% Design data:
% X is a matrix with features (columns)
% d is the id... |
github | domingomery/Balu-master | Bcl_boosting.m | .m | Balu-master/Classification/Bcl_boosting.m | 3,635 | utf_8 | a493572ca995c53735727792a4c6346c | % ds = Bcl_boosting(X,d,Xt,options) Training & Testing together
% options = Bcl_boosting(X,d,options) Training only
% ds = Bcl_boosting(Xt,options) Testing only
%
% Toolbox: Balu
% Boosting classifier.
%
% Design data:
% X is a matrix with features (columns)
% d is the ide... |
github | domingomery/Balu-master | Bcl_BalanceCascade.m | .m | Balu-master/Classification/Bcl_BalanceCascade.m | 5,067 | utf_8 | 93da7b82d0353300ef25dee1bd45b5d8 | % ds = Bcl_BalanceCascade(X,d,Xt,options) Training & Testing together
% options = Bcl_BalanceCascade(X,d,options) Training only
% ds = Bcl_BalanceCascade(Xt,options) Testing only
%
% Toolbox: Balu
%
% BalanceCascade classifier for imabalance data, where the label for
% the majority cla... |
github | domingomery/Balu-master | Bcl_lda.m | .m | Balu-master/Classification/Bcl_lda.m | 3,563 | utf_8 | 1d8a24cda0b09566621fef8958f99f58 | % ds = Bcl_lda(X,d,Xt,[]) Training & Testing together
% options = Bcl_lda(X,d,[]) Training only
% ds = Bcl_lda(Xt,options) Testing only
%
% Toolbox: Balu
% LDA (linear discriminant analysis) classifier.
% We assume that the classes have a common covariance matrix
%
% Design data:
% ... |
github | domingomery/Balu-master | Bcl_svmplus.m | .m | Balu-master/Classification/Bcl_svmplus.m | 3,234 | utf_8 | bb7d3750d3dbe492a9a2f4f1e781e792 | % ds = Bcl_svmplus(X,d,Xt,options) Training & Testing together
% options = Bcl_svmplus(X,d,options) Training only
% ds = Bcl_svmplus(Xt,options) Testing only
%
% Toolbox: Balu
% Classifier using Support Vector Machine approach using Bioinformatics
% Toolbox of Matlab using tree algorithm when ... |
github | domingomery/Balu-master | Bcl_dcs.m | .m | Balu-master/Classification/Bcl_dcs.m | 3,784 | utf_8 | a69377e4865b0a7d5dab93ec732c4e6e | % ds = Bcl_dcs(X,d,Xt,options) Training & Testing together
% options = Bcl_dcs(X,d,options) Training only
% ds = Bcl_dcs(Xt,options) Testing only
%
% Toolbox: Balu
% Dynamic classifier selection based on multiple
% classifier behaviour after Giacinto (2001).
%
% Design data:
% ... |
github | domingomery/Balu-master | Bcl_dmin.m | .m | Balu-master/Classification/Bcl_dmin.m | 2,329 | utf_8 | b9e56b36e6824dbb8f2d8a10440aa3c2 | % ds = Bcl_dmin(X,d,Xt,[]) Training & Testing together
% options = Bcl_dmin(X,d,[]) Training only
% ds = Bcl_dmin(Xt,options) Testing only
%
% Toolbox: Balu
% Classifier using Euclidean minimal distance
%
% Design data:
% X is a matrix with features (columns)
% d is the ideal c... |
github | domingomery/Balu-master | Bcl_boostVJ.m | .m | Balu-master/Classification/Bcl_boostVJ.m | 3,887 | utf_8 | 85651554828473f7278508b83f8464ed | % ds = Bcl_boostVJ(X,d,Xt,options) Training & Testing together
% options = Bcl_boostVJ(X,d,options) Training only
% ds = Bcl_boostVJ(Xt,options) Testing only
%
% Toolbox: Balu
% Boosting algorithm after Viola Jones. It uses only one feature
% per weak classifier.
%
% Design data:
% X ... |
github | domingomery/Balu-master | Bcl_svm.m | .m | Balu-master/Classification/Bcl_svm.m | 2,342 | utf_8 | 1dc0cca34cf006d5f9b2c464b841fe37 | % ds = Bcl_svm(X,d,Xt,options) Training & Testing together
% options = Bcl_svm(X,d,options) Training only
% ds = Bcl_svm(Xt,options) Testing only
%
% Toolbox: Balu
% Support Vector Machine approach using the Bioinformatics Toolbox.
%
% Design data:
% X is a matrix with features (columns)... |
github | domingomery/Balu-master | Bcl_EasyEnsemble.m | .m | Balu-master/Classification/Bcl_EasyEnsemble.m | 4,569 | utf_8 | 8005b400cf9893ba6136fc41ee4ea432 | % ds = Bcl_EasyEnsemble(X,d,Xt,options) Training & Testing together
% options = Bcl_EasyEnsemble(X,d,options) Training only
% ds = Bcl_EasyEnsemble(Xt,options) Testing only
%
% Toolbox: Balu
%
% EasyEnsemble classifier for imabalance data, where the label for
% the majority class is 0 ... |
github | domingomery/Balu-master | Bcl_RandomForest.m | .m | Balu-master/Classification/Bcl_RandomForest.m | 2,124 | utf_8 | 1096f36a25d40e6086f5ed1e348346cd | % ds = Bcl_RandomForest(X,d,Xt,[]) Training & Testing together
% options = Bcl_RandomForest(X,d,[]) Training only
% ds = Bcl_RandomForest(Xt,options) Testing only
%
% Toolbox: Balu
% Classifier using Random Forest. This implementation uses command
% TreeBagger of Statistics and Machine Learni... |
github | domingomery/Balu-master | Bcl_pnn.m | .m | Balu-master/Classification/Bcl_pnn.m | 2,086 | utf_8 | 2010fa63b6311a4d726f529314f4a5f9 | % ds = Bcl_pnn(X,d,Xt,options) Training & Testing together
% options = Bcl_pnn(X,d,options) Training only
% ds = Bcl_pnn(Xt,options) Testing only
%
% Toolbox: Balu
% Probabilistic neural network (Neural Network Toolbox required).
%
% Design data:
% X is a matrix with features (co... |
github | domingomery/Balu-master | Bcl_det22.m | .m | Balu-master/Classification/Bcl_det22.m | 4,367 | utf_8 | c4bdbb211abe33495dceb60d0977652e | % ds = Bcl_det22(X,d,Xt,[]) Training & Testing together
% options = Bcl_det22(X,d,[]) Training only
% ds = Bcl_det22(Xt,options) Testing only
%
% Toolbox: Balu
% Quadratic Detector Design for ONLY two classes and two features
%
% Design data:
% X is a matrix with features (columns)
%... |
github | domingomery/Balu-master | Bcl_gui2.m | .m | Balu-master/Classification/Bcl_gui2.m | 36,876 | utf_8 | 129b93f09dc609ac7917c0a4c62a099f | % Bcl_gui2
%
% Toolbox: Balu
%
% Graphic User Interface for feature extraction.
%
% (c) GRIMA-DCCUC, 2011
% http://grima.ing.puc.cl
function varargout = Bcl_gui2(varargin)
% BCL_GUI2 M-file for Bcl_gui2.fig
% BCL_GUI2, by itself, creates a new BCL_GUI2 or raises the existing
% singleton*.
%
% H = BCL... |
github | domingomery/Balu-master | Bcl_nbnnxi.m | .m | Balu-master/Classification/Bcl_nbnnxi.m | 1,759 | utf_8 | f255c05dc65a3eacd2c5dcc9ff666156 | % ds = Bcl_nbnnxi(X,d,Xt,D);
%
% Toolbox: Balu
% Naive Bayes Nearest Neighbor for histograms using Xi distance
%
% Design data:
% X is the feature matrix having M histograms of D bins each per sample
% d is the ideal classification for X
%
% Test data:
% Xt is the feature matrix
%
% options.D: num... |
github | domingomery/Balu-master | Bcl_SMOTEBoost.m | .m | Balu-master/Classification/Bcl_SMOTEBoost.m | 8,503 | utf_8 | 66c5df0906bf8d2baf2fd6db742f6c8e | % ds = Bcl_SMOTEBoost(X,d,Xt,options) Training & Testing together
% options = Bcl_SMOTEBoost(X,d,options) Training only
% ds = Bcl_SMOTEBoost(Xt,options) Testing only
%
% Toolbox: Balu
% SMOTEBoost classifier for imbalance data, where the label for
% the majority class is 0 and the labe... |
github | domingomery/Balu-master | Bcl_structure.m | .m | Balu-master/Classification/Bcl_structure.m | 4,155 | utf_8 | 752e75839ca633932bb42a627cd146c4 | % ds = Bcl_structure(X,d,Xt,options) Training & Testing together
% options = Bcl_structure(X,d,options) Training only
% ds = Bcl_structure(Xt,options) Testing only
%
% Toolbox: Balu
% Classification using Balu classifier(s) defined in structure b.
%
% Design data:
% X is a matrix with fe... |
github | domingomery/Balu-master | Bcl_knn.m | .m | Balu-master/Classification/Bcl_knn.m | 2,658 | utf_8 | a6ecb903c356122f3528e31bbc0a7501 | % ds = Bcl_knn(X,d,Xt,options) Training & Testing together
% options = Bcl_knn(X,d,options) Training only
% ds = Bcl_knn(Xt,options) Testing only
%
% Toolbox: Balu
% KNN (k-nearest neighbors) classifier using randomized kd-tree
% forest from FLANN. This implementation requires VLFeat Too... |
github | domingomery/Balu-master | Bcl_det21.m | .m | Balu-master/Classification/Bcl_det21.m | 5,393 | utf_8 | 1b3900ebaafed895dce91437f10e8dfd | % ds = Bcl_det21(X,d,Xt,[]) Training & Testing together
% options = Bcl_det21(X,d,[]) Training only
% ds = Bcl_det21(Xt,options) Testing only
%
% Toolbox: Balu
% Linear Detector Design for ONLY two classes and two features
%
% Design data:
% X is a matrix with features (columns)
% ... |
github | domingomery/Balu-master | Bcl_bagging.m | .m | Balu-master/Classification/Bcl_bagging.m | 2,933 | utf_8 | 67cf8b329006d68067f14b35c6931f7d | % ds = Bcl_bagging(X,d,Xt,options) Training & Testing together
% options = Bcl_bagging(X,d,options) Training only
% ds = Bcl_bagging(Xt,options) Testing only
%
% Toolbox: Balu
% Bagging classifier.
%
% Design data:
% X is a matrix with features (columns)
% d is the ideal c... |
github | domingomery/Balu-master | Bcl_balu.m | .m | Balu-master/Classification/Bcl_balu.m | 8,762 | utf_8 | 08de931e4051bf051d0bb7f08aa4bdce | % [bcs,selec,sp] = Bcl_balu(X,d,bcl,bfs,options)
%
% Toolbox: Balu
% Feature and classifier selection tool.
% Exhaustive search of the best classifier of the classifiers given in
% bcl structure using the features selected by feature selection
% algorithms given in bfs structure.
%
% X features
% d id... |
github | domingomery/Balu-master | Bcl_gui.m | .m | Balu-master/Classification/Bcl_gui.m | 36,853 | utf_8 | e6b9ad20f20936a05c1bc78b6f5d41ba | % Bcl_gui
%
% Toolbox: Balu
%
% Graphic User Interface for feature extraction.
%
% (c) GRIMA-DCCUC, 2011
% http://grima.ing.puc.cl
function varargout = Bcl_gui(varargin)
% BCL_GUI M-file for Bcl_gui.fig
% BCL_GUI, by itself, creates a new BCL_GUI or raises the existing
% singleton*.
%
% H = BCL_GUI r... |
github | domingomery/Balu-master | Bcl_maha.m | .m | Balu-master/Classification/Bcl_maha.m | 2,698 | utf_8 | 43d317366ec54fb5636dc6667b5b64eb | % ds = Bcl_maha(X,d,Xt,[]) Training & Testing together
% options = Bcl_maha(X,d,[]) Training only
% ds = Bcl_maha(Xt,options) Testing only
%
% Toolbox: Balu
% Classifier using Mahalanobis minimal distance
%
% Design data:
% X is a matrix with features (columns)
% d is the ideal... |
github | domingomery/Balu-master | Bcl_ann.m | .m | Balu-master/Classification/Bcl_ann.m | 2,237 | utf_8 | b0f661155cf6ef4935a12643ecc8cfe9 | % ds = Bcl_ann(X,d,Xt,[]) Training & Testing together
% options = Bcl_ann(X,d,[]) Training only
% ds = Bcl_ann(Xt,options) Testing only
%
% Toolbox: Balu
% Simple Neural Network using Neural Network Toolbox of Matlab using
% softmax
%
% Design data:
% X is a matrix with features (columns)
... |
github | domingomery/Balu-master | Bcl_exe.m | .m | Balu-master/Classification/Bcl_exe.m | 1,858 | utf_8 | 0b2d313ca8e130505178f8c3e71b941d | % ds = Bcl_exe(bname,X,d,Xt,options) Training & Testing together
% options = Bcl_exe(bname,X,d,options) Training only
% ds = Bcl_exe(bname,Xt,options) Testing only
%
% Toolbox: Balu
% Classification using Balu classifier bname and options.
%
% Design data:
% bname can be any name of a Ba... |
github | domingomery/Balu-master | Bcl_construct.m | .m | Balu-master/Classification/Bcl_construct.m | 1,260 | utf_8 | ee3c272415b3216d2e8de44f29ea4657 | % This function is not a classifier!!!
% This function is called by Balu classifier functions (such as Bcl_lda) to
% build the training and testing data.
function [train,test,X,d,Xt,options] = Bcl_construct(varargin)
train = 0;
test = 0;
switch nargin
case 2 % [ds,options] = Bcl_svm(Xt,options) % testing ... |
github | domingomery/Balu-master | Bcl_nnglm.m | .m | Balu-master/Classification/Bcl_nnglm.m | 29,935 | utf_8 | 4940b543e9d6fbf5fc6fd16c1c12999c | % ds = Bcl_nnglm(X,d,Xt,[]) Training & Testing together
% options = Bcl_nnglm(X,d,[]) Training only
% ds = Bcl_nnglm(Xt,options) Testing only
%
% Toolbox: Balu
% Neural Network using a Generalized Linear Model
%
% Design data:
% X is a matrix with features (columns)
% d is the ideal cla... |
github | domingomery/Balu-master | Bcl_pegasos.m | .m | Balu-master/Classification/Bcl_pegasos.m | 2,930 | utf_8 | c91b620ac1c40013c50c4e10b5daa6ea | % ds = Bcl_pegasos(X,d,Xt,[]) Training & Testing together
% options = Bcl_pegasos(X,d,[]) Training only
% ds = Bcl_pegasos(Xt,options) Testing only
%
% Toolbox: Balu
% Classifier using Pegasos Support Vector Machine approach using
% VLFeat Toolbox of Matlab.
%
% Design data:
% X is a matri... |
github | domingomery/Balu-master | Bcl_knn_old.m | .m | Balu-master/Classification/Bcl_knn_old.m | 2,296 | utf_8 | 44a58443cff88610aa16647e0fd21aa6 | % ds = Bcl_knn_old(X,d,Xt,options) Training & Testing together
% options = Bcl_knn_old(X,d,options) Training only
% ds = Bcl_knn_old(Xt,options) Testing only
%
% Toolbox: Balu
% KNN (k-nearest neighbors) classifier. This implementation does not
% require VLFeat Toolbox. If you have it, ... |
github | domingomery/Balu-master | Bcl_qda.m | .m | Balu-master/Classification/Bcl_qda.m | 3,512 | utf_8 | 72793a9bcf956be02ed7064d37f3042e | % ds = Bcl_qda(X,d,Xt,[]) Training & Testing together
% options = Bcl_qda(X,d,[]) Training only
% ds = Bcl_qda(Xt,options) Testing only
%
% Toolbox: Balu
% QDA (quadratic discriminant analysis) classifier.
%
% Design data:
% X is a matrix with features (columns)
% d is the idea... |
github | domingomery/Balu-master | Bhelp.m | .m | Balu-master/Help/Bhelp.m | 707 | utf_8 | f0c29178016bd491d646a1bfb02ab86a | % Toolbox: Balu
% Display help text for Balu Matlab Toolbox
%
% D.Mery, PUC-DCC, Jun 2010
% http://dmery.ing.puc.cl
%
function Bhelp(t)
disp('Help for Balu Matlab Toolbox - (c) GRIMA, PUC-DCC')
disp(' ')
if exist('t','var')
help BhelpImageProcessing
help BhelpFeatureExtraction
help BhelpFeatureTransformation
h... |
github | domingomery/Balu-master | Btr_sfm.m | .m | Balu-master/Tracking/Btr_sfm.m | 3,816 | utf_8 | d5197ff95ed6bc65dbfcbf6db7ad98ec | % [P,H1] = Btr_sfm(kp,Ho,options)
%
% Toolbox: Balu
%
% Structure from Motion.
%
% kp keypoints structure according function Bsq_des (see help)
%
% Ho is a matching multi-views matrix with Nxn indices for N matchings
% in n views.
%
% options.sfs_method = 1 for affine projection and 2 for projective
% ... |
github | domingomery/Balu-master | Btr_detection.m | .m | Balu-master/Tracking/Btr_detection.m | 3,430 | utf_8 | 1855fe71e62a2b3a8bc61fcbe502173d | % [Tf,kp1,kp2,f,P] = Btr_detection(f,op1,op2,op3)
%
% Toolbox: Balu
%
% Detection by tracking.
% Fine Detection in Rigid Objects using Multi-views
%
% This method consists of two steps: Structure Estimation, to obtain a
% geometric model of the multi-views from the object itself, and
% Details Detection,... |
github | domingomery/Balu-master | Btr_sift2.m | .m | Balu-master/Tracking/Btr_sift2.m | 3,329 | utf_8 | 85024832d1f84c1b18c005fbb25e7e0d | % Bpq = Btr_sift2(kp,p,q,options)
%
% Toolbox: Balu
%
% Matching points between views p and image q using SIFT keypoints.
%
% kp keypoints structure according function Bsq_des (see help)
% options.matching 1: matching is estimated directly using vl_ubcmatch
% function.
% options.matching 2: matching is e... |
github | domingomery/Balu-master | Btr_analysis.m | .m | Balu-master/Tracking/Btr_analysis.m | 7,186 | utf_8 | 884c4d922c980fd27cc8adee696fe0e1 | % [Z,f] = Btr_analysis(kp,Y,P,files,options)
%
%
% Toolbox: Balu
%
% Track analysis. Btr_analysis selects those trajectories that
%
% The 3D reconstructed points are reprojected in those views where the
% segmentation may have failed to obtain the complete track in all
% views. The reprojected points should... |
github | domingomery/Balu-master | Btr_gui.m | .m | Balu-master/Tracking/Btr_gui.m | 21,075 | utf_8 | f0c7868ae2d55491ee20072bfd37811e | function varargout = Btr_gui(varargin)
% BTR_GUI M-file for Btr_gui.fig
% BTR_GUI, by itself, creates a new BTR_GUI or raises the existing
% singleton*.
%
% H = BTR_GUI returns the handle to a new BTR_GUI or the handle to
% the existing singleton*.
%
% BTR_GUI('CALLBACK',hObject,eventData,handl... |
github | domingomery/Balu-master | Btr_3.m | .m | Balu-master/Tracking/Btr_3.m | 2,370 | utf_8 | f3015dd26a50808ddfd883515f1cd455 | % C = Btr_3(kp,B,T,options)
%
% Toolbox: Balu
%
% Matching points between all three views p, q and r of a sequence, for
% p=1:n-1, and for q=p+1:p+m (n is the number of views in the sequence
% and m is defined by options.mviews)
%
% kp keypoints structure according function Bsq_des (see help)
%
% B is a ... |
github | domingomery/Balu-master | Btr_2.m | .m | Balu-master/Tracking/Btr_2.m | 2,406 | utf_8 | 8ab5e275d289a73d541711ebcb486200 | % B = Btr_2(kp,F,options)
%
% Toolbox: Balu
%
% Matching points between all two views p and q of a sequence, for
% p=1:n-1, and for q=p+1:p+m (n is the number of views in the sequence
% and m is defined by options.mviews)
%
% kp keypoints structure according function Bsq_des (see help)
%
% F are the fund... |
github | domingomery/Balu-master | Btr_merge.m | .m | Balu-master/Tracking/Btr_merge.m | 1,395 | utf_8 | 6e478daeb63770141d0fe68f628f06fa | % E = Btr_merge(kp,D)
%
% Toolbox: Balu
%
% Merge tracks with common matching points.
%
% kp keypoints structure according function Bsq_des (see help)
%
% D is a Nxm matrix with N matchings in m views. The output E is a
% matrix with merged trajectories with common keypoints.
%
% Example:
% See exampl... |
github | domingomery/Balu-master | Btr_siftn.m | .m | Balu-master/Tracking/Btr_siftn.m | 2,227 | utf_8 | 642eb185c2713873c1860f245fa1909e | % Bo = Btr_siftn(kp,options)
%
% Toolbox: Balu
%
% Matching points in all two consecutive views of a sequence using SIFT
% keypoints.
%
% kp keypoints structure according function Bsq_des (see help)
% options.matching 1: matching is estimated using vl_ubcmatch
% function only.
% options.matching 2: m... |
github | domingomery/Balu-master | Btr_sfseq.m | .m | Balu-master/Tracking/Btr_sfseq.m | 3,571 | utf_8 | 47d246d25221fce477d8280a45457137 | % [P,f,kp,H1] = Btr_sfseq(f,options)
%
% Toolbox: Balu
%
% Structure from an image sequence
%
% The structure estimation is obtained by computing a geometric model of
% the multi-views from the object itself. The geometric model is
% estimated by a bundle adjustment algorithm on stable SIFT keypoints
% a... |
github | domingomery/Balu-master | Btr_join.m | .m | Balu-master/Tracking/Btr_join.m | 2,058 | utf_8 | dff038a9c6438326c1969408fbb625cb | % H2 = Btr_join(H0,H1,options)
% H2 = Btr_join(H0,[],options)
% H2 = Btr_join(H0,p,options)
%
% Toolbox: Balu
%
% Join of matching points.
%
% Hk (k=0,1,2) is a matching multi-views matrix with Nkxnk indices for
% Nk matchings in nk views. H2 has the tracks that have p common
% elements in H1 and H0 (the la... |
github | domingomery/Balu-master | Btr_windows.m | .m | Balu-master/Tracking/Btr_windows.m | 4,161 | utf_8 | 4962979d095ce0025cac9175fc689e95 | % [Z,f] = Btr_analysis(kp,Y,P,options)
%
%
% Toolbox: Balu
%
% Track analysis. Btr_analysis selects those trajectories that
%
% The 3D reconstructed points are reprojected in those views where the
% segmentation may have failed to obtain the complete track in all
% views. The reprojected points should corre... |
github | domingomery/Balu-master | Btr_classify.m | .m | Balu-master/Tracking/Btr_classify.m | 2,745 | utf_8 | f3bcdb7fbc472f153f12f52c2860cb89 | % [Z,f] = Btr_analysis(W,Iw,options)
%
%
% Toolbox: Balu
%
% Track analysis. Btr_analysis selects those trajectories that
%
% The 3D reconstructed points are reprojected in those views where the
% segmentation may have failed to obtain the complete track in all
% views. The reprojected points should corresp... |
github | domingomery/Balu-master | Btr_plot.m | .m | Balu-master/Tracking/Btr_plot.m | 3,602 | utf_8 | fb6b28fefb9a82d93393afb63688dc30 | % Btr_plot(kp,A,files,options)
%
% Toolbox: Balu
%
% Plot of tracks.
%
% kp keypoints structure according function Bsq_des (see help)
%
% A is a Nxm matrix with N matchings in m views. Each row is a track to be plot.
%
% files is a structure that define the images of the sequence according
% to function ... |
github | domingomery/Balu-master | Bev_holdout.m | .m | Balu-master/PerformanceEvaluation/Bev_holdout.m | 2,845 | utf_8 | 7fb64f214210b186a84fd4c311dd2ae2 | % [T,p] = Bev_holdout(X,d,options)
%
% Toolbox: Balu
% Holdout evaluation of a classifier.
%
% X is a matrix with features (columns)
% d is the ideal classification for X
%
% options.b is a Balu classifier or several classifiers (see example)
% options.s is the portion of data used for training,... |
github | domingomery/Balu-master | Bev_crossval_old.m | .m | Balu-master/PerformanceEvaluation/Bev_crossval_old.m | 5,770 | utf_8 | fa643680465aafd48ff665f4df2684ba | % [p,ci] = Bev_crossval(X,d,options)
%
% Toolbox: Balu
%
% Cross-validation evaluation of a classifier.
%
% v-fold Cross Validation in v groups of samples X and classification d
% according to given method. If v is equal to the number of samples,
% i.e., v = size(X,1), this method works as the origi... |
github | domingomery/Balu-master | Bev_crossval.m | .m | Balu-master/PerformanceEvaluation/Bev_crossval.m | 5,997 | utf_8 | 68a413ddd76d351a1746720e0b67c908 | % [p,ci] = Bev_crossval(X,d,options)
%
% Toolbox: Balu
%
% Cross-validation evaluation of a classifier.
%
% v-fold Cross Validation in v groups of samples X and classification d
% according to given method. If v is equal to the number of samples,
% i.e., v = size(X,1), this method works as the origi... |
github | domingomery/Balu-master | Bev_reclassification.m | .m | Balu-master/PerformanceEvaluation/Bev_reclassification.m | 2,644 | utf_8 | d34935d5b341fadfe0314ea080c119cb | % [T,p] = Bev_holdout(X,d,options)
%
% Toolbox: Balu
% Holdout evaluation of a classifier.
%
% X is a matrix with features (columns)
% d is the ideal classification for X
%
% options.b is a Balu classifier or several classifiers (see example)
% options.s is the portion of data used for training,... |
github | domingomery/Balu-master | Bev_performance.m | .m | Balu-master/PerformanceEvaluation/Bev_performance.m | 1,953 | utf_8 | 1de31e573a11374b68e74af5f8c1147a | % p = Bev_erformance(d1,d2,nn)
%
% Toolbox: Balu
% Performance evaluation between two classifications, e.g., ideal (d1)
% and real (d2) classification.
%
% d1 and d2 are vectors or matrices (vector Nxn1 and Nxn2 respectivelly)
% at least n1 or n2 muts be one. N is the number of samples.
% p is t... |
github | domingomery/Balu-master | Bev_bootstrap0632.m | .m | Balu-master/PerformanceEvaluation/Bev_bootstrap0632.m | 3,392 | utf_8 | 400d0c9ec461af76cb33a31add331f5c | % [p,ci] = Bev_bootstrap0632(X,d,options)
%
% Toolbox: Balu
% 0.632 Bootstrap in B bootstrap samples of X and classification d
% according to given classifier.
%
% X is a matrix with features (columns)
% d is the ideal classification for X
%
% options.b is a Balu classifier or several classifier... |
github | domingomery/Balu-master | Bev_jackknife.m | .m | Balu-master/PerformanceEvaluation/Bev_jackknife.m | 2,509 | utf_8 | b6e2b4239fea73942807f04cbd9e15f8 | % [T,p] = Bev_jackknife(X,d,options)
%
% Toolbox: Balu
% Holdout evaluation of a classifier.
%
% v-fold Cross Validation in v groups of samples X, where v is the
% number of sanples (i.e., v = size(X,1)). The training will be in X
% without sample i and testing in sample i. ci is the confidence inte... |
github | domingomery/Balu-master | Bev_bootstrap.m | .m | Balu-master/PerformanceEvaluation/Bev_bootstrap.m | 3,209 | utf_8 | f011961c0a427d0fab35ff5e7d3156de | % [p,ci] = Bev_bootstrap(X,d,options)
%
% Toolbox: Balu
% Bootstrap evaluation in B bootstrap samples of X and classification d
% according to given classifier.
%
% X is a matrix with features (columns)
% d is the ideal classification for X
%
% options.b is a Balu classifier or several classifie... |
github | domingomery/Balu-master | Bev_confusion.m | .m | Balu-master/PerformanceEvaluation/Bev_confusion.m | 1,042 | utf_8 | ecf6830edc7f91e3b448d3f15a5296b4 | % function [T,p] = Bev_confusion(d,ds,nn);
%
% Toolbox: Balu
% Confusion Matrix and Performance of a classification
%
% d is the ideal classification (vector Nx1 with N samples)
% ds is the classified data (vector Nx1)
% T is the confusion matrix (nxn) for n classes
% T(i,j) indicates the number ... |
github | domingomery/Balu-master | Bev_roc.m | .m | Balu-master/PerformanceEvaluation/Bev_roc.m | 2,854 | utf_8 | 3be5432a835e96264b410b715b58e387 | % [Az,Sn,Sp1,t] = Bev_roc(z,d,show)
%
% Toolbox: Balu
% ROC Analysis for feature z with classification c. show = 1 indicates
% that the ROC curve will be displayed.
% Az is the area under the ROC curve
% Sn and Sp1 are the coordinates of Sensitibity and 1-Specificity of
% optimal point of ROC cur... |
github | domingomery/Balu-master | Bft_uninorm.m | .m | Balu-master/FeatureTransformation/Bft_uninorm.m | 373 | utf_8 | f9c849160e8c3b97346ccb03f31f1176 | % Xnew = Bft_uninorm(X)
%
% Toolbox: Balu
%
% Normalization of features X: each row of Xnew has norm = 1
%
% Example:
% load datareal
% Xnew = Bft_uninorm(f);
%
% (c) Grima, PUC-DCC, 2013: D. Mery
% http://dmery.ing.puc.cl
function Xnew = Bft_uninorm(X)
[N,M] = size(X);
Xnew = zeros(N,M);
f... |
github | domingomery/Balu-master | Bft_lseft.m | .m | Balu-master/FeatureTransformation/Bft_lseft.m | 2,109 | utf_8 | 4dbb32d0a819ca70760367b31194b276 | % [Y,selec,th] = Bft_lseft(X,d,options)
%
% Toolbox: Balu
% Feature transformacion using LSE-forward algorithm
%
% input: X feature matrix
% options.m number of features to be selected
% optoins.show = 1 displays results
% options.pca = 1 for PCA and = 0 for PLS
%
% outp... |
github | domingomery/Balu-master | Bft_plsr.m | .m | Balu-master/FeatureTransformation/Bft_plsr.m | 3,192 | utf_8 | 3a82fc389363f7d89009573791119cf9 | % [T,U,P,Q,W,B] = Bft_plsr(X,d,options)
%
% Toolbox: Balu
% Feature transformation using Partial Least Squares Regression with
% NIPALS algorithm.
% X: Input matrix with features
% d: Vector with ideal classifcation.
% m: Number of principal components to be selected.
% T: Loadings of X (m tra... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.